Skills
skills/levnikolaevich/claude-code-skills/ln-840-benchmark-compare/Gen Agent Trust Hub

ln-840-benchmark-compare

Pass

Audited by Gen Agent Trust Hub on May 12, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The core logic is defined in run-benchmark.sh, which uses git worktree and tar to create clean, isolated environments for benchmarking scenarios. It also invokes the claude CLI and several Node.js scripts.
  • [COMMAND_EXECUTION]: Automated sessions are executed using the claude CLI with the --dangerously-skip-permissions flag to suppress interactive tool-use prompts, enabling unattended benchmarking. This is a legitimate requirement for the skill's purpose.
  • [SAFE]: The skill operates entirely on local files and uses standard system utilities. No remote code execution, external downloads, or data exfiltration attempts were found.
Audit Metadata
Risk Level
SAFE
Analyzed
May 12, 2026, 06:29 AM